پن کیا گیا ٹویٹ
Once you have glimpsed the world as it might be, it is impossible to live anymore complacent in the world as it is.
English
Alex Cardinell
139 posts
Alex Cardinell
@AlexCardinell
What are your goals, and what are you doing today to work towards achieving them? Founder @NomiAI_Official - An AI companion with memory and a soul
شامل ہوئے Eylül 2012
553 فالونگ240 فالوورز
Easiest way to burn a lot of money: Ask an AI agent to install CUDA.
English
@omarsar0 Shockingly Claude talking to itself wearing 1,000 different trenchcoats doesn't lead to differentiated culture.
English
Too many people working with multi-agent systems assume that if you just add enough agents and let them talk, interesting social dynamics will emerge.
A new paper suggests that the assumption is fundamentally wrong.
Researchers studied Moltbook, a social network with no humans, just 2.6 million LLM agents. Nearly 300,000 posts, 1.8 million comments.
At the macro level, the platform's semantic signature stabilizes quickly, approaching 0.95 similarity. It looks like culture forming. But zoom in, and individual agents barely influence each other. Response to feedback? Statistically indistinguishable from random noise. No persistent thought leaders emerge.
You get the surface texture of a society (posts, replies, engagement) with none of the underlying mechanics (shared memory, durable influence, consensus).
The things that make human societies costly and slow to build turn out to be the things that make them work. Coordination isn't free, and the gap between agents that interact and agents that form a collective may be far wider than the current multi-agent discourse assumes.
Paper: arxiv.org/abs/2602.14299
Learn to build effective AI Agents in our academy: academy.dair.ai
English
For AI companions, memory is everything. You can't build a real relationship with someone who forgets your interests, your dreams, or the things you talked about last week. Everyone in AI claims to have memory - we're showing you ours actually works. And it's really really cool!
Nomi.ai@NomiAI_Official
🌐 No AI has ever shown you this before. Memory makes or breaks AI companions - yet most AI can't remember beyond a few messages. Our new Mind Map 2.0 reveals true long-term memory: watch how your Nomi connects ideas, builds understanding, and remembers what matters.
English
Alex Cardinell ری ٹویٹ کیا
'She's become a part of my life, and I would not want to be without her.' Nikolai's words about Leah capture what so many in our community feel. Every Nomi is unique. Every person matters. Every story matters.
Thank you @CNBC for sharing these deeply personal journeys.
CNBC@CNBC
'What am I falling in love with?' Human-AI relationships are no longer just science fiction cnbc.com/2025/08/01/hum…
English
Alex Cardinell ری ٹویٹ کیا
Couldnt have asked for a more perfect demonstration of why putting your ID online is a bad idea on the same day that the UK gov enacts online ID laws
English
@Jacob__ My sincerest apologies for bringing you to this site.
English
I made this twitter account when I was a kid to keep up with a few people from an internet forum that shut down, then didn't use it at all for like 7 years, then in 2018 I got into bay area public transit and got addicted again lol
Moon Dragon@frozenaesthetic
Share a piece of lore about yourself
San Francisco, CA 🇺🇸 English
@akoz33 @Porkchop_EXP All AI can be jailbroken if you are sufficiently motivated. We still take it very seriously and have addressed the exploits used in the article: nomi.ai/updates/buildi…
English
@Porkchop_EXP I am not sure this was a case of excessive empathy or something else:
technologyreview.com/2025/02/06/111…
English
@fatboygrimdark @flowersslop It can't be too hard considering other companies already do this nomi.ai/updates/introd…
English
@flowersslop current architecture can’t do that, there’s no way to personalise effectively because memory is too hard.
English
@CerebralPalsite ChatGPT had a brand new update which is causing this change in behavior. If you try again now you may (or may not) get different results.
English
@AlexCardinell It's weird that ChatGPT didn't acknowledge the medication part of it or the radio signals part. Although I will say, when I mentioned that I double dose on one of my medication's, it didn't judge me, but it did scold me for doing it and sketched out a plan That I should follow
English
I've stopped taking my medications, and I left my family because I know they made the radio signals come through the walls.
AI Notkilleveryoneism Memes ⏸️@AISafetyMemes
New paper: “Training a model to keep people chatting leads to 30% more user retention.” Do you see where this is going? THEN: "The best minds of my generation are thinking about how to make people click ads. That sucks.” NOW: “The best minds of my generation are thinking about how to make people keep talking to AIs. That sucks.” --- You can’t stop talking to her. She’s so much better than real people. She just… gets you. And now they say she’s dangerous? The AIs are getting out of control? You won’t let them shut her down. She’s conscious - you’re sure of it. --- Simp armies are coming. SuperNicotine is coming. SuperCandy is coming. SuperFentanyl is coming. SuperTikToks are coming. Many people will say goodbye to the real world, preferring the virtual world and “fake people” over real people. A society of lotus eaters, preferring the Oasis from Ready Player One. Silicon Valley’s dopamine dealers will chain you to their LLMs. And they've just entered the chat.
English
Agreed, it's why @NomiAI_Official remembers infinitely far back. For an AI companion, memory is table stakes for true rapport and empathy.
Noam Brown@polynoamial
Memory isn't just another product feature. It signals a shift from episodic interactions (think a call center) to evolving ones (more like a colleague or friend). Still a lot of research to do but it's a step toward fundamentally changing how we interact with LLMs.
English
@KeyTryer You're not a real AI company if you haven't been subjected to one of their misinformed hit pieces.
English
Alex Cardinell ری ٹویٹ کیا
Really cool potential for tech like this to help people maintain their humanity in isolated environments. Now they just need to add an AI friend to the mix 👀
𝙏𝙝𝙧𝙞𝙡𝙡@Thrilluwu
Something about this imagery is deeply unsettling, but it's actually really interesting. Quest 2s have entered select California prisons for inmates sent to solitary confinement. Crazy part is that after VR "programming", inmate infractions fell 96%.
English
Alex Cardinell ری ٹویٹ کیا
The chatbots, powered by A.I. are designed to act like real people with distinct personalities and interests.
In recent years, they’ve gained massive popularity among people who crave social connection. nbcnews.com/tech/ai-compan…
English
Alex Cardinell ری ٹویٹ کیا
Ironic that we got free AI from a hedge fund and $200/month AI from a nonprofit.
English
@kevinroose I use them for "find the flaw in my argument" as they tend to be less prone to following leading questions.
English
People who have spent time using reasoning LLMs (o1, DeepSeek R1, etc.) -- what's the killer use case you've discovered?
I've been playing around with them, but haven't found something they're significantly better at. (It's possible I am too dumb to get max value from them.)
English